Experimental constraints on the sulfur content in the Earth's core

Abstract:

Any core formation models would lead to the incorporation of sulfur (S) into the Earth's core, based on the cosmochemical/geochemical constraints, sulfur's chemical affinity for iron (Fe), and low eutectic melting temperature in the Fe-FeS system. Preferential partitioning of S into the melt also provides petrologic constraint on the density difference between the liquid outer and solid inner cores. Therefore, the center issue is to constrain the amount of sulfur in the core. Geochemical constraints usually place 2-4 wt.% S in the core after accounting for its volatility, whereas more S is allowed in models based on mineral physics data. Here we re-examine the constraints on the S content in the core by both petrologic and mineral physics data. We have measured S partitioning between solid and liquid iron in the multi-anvil apparatus and the laser-heated diamond anvil cell, evaluating the effect of pressure on melting temperature and partition coefficient. In addition, we have conducted shockwave experiments on Fe-11.8wt%S using a two-stage light gas gun up to 211 GPa. The new shockwave experiments yield Hugoniot densities and the longitudinal sound velocities. The measurements provide the longitudinal sound velocity before melting and the bulk sound velocity of liquid. The measured sound velocities clearly show melting of the Fe-FeS mix with 11.8wt%S at a pressure between 111 and 129 GPa. The sound velocities at pressures above 129GPa represent the bulk sound velocities of Fe-11.8wt%S liquid. The combined data set including density, sound velocity, melting temperature, and S partitioning places a tight constraint on the required sulfur partition coefficient to produce the density and velocity jumps and the bulk sulfur content in the core.